When a client hands over photographs, videos, recordings, or documents for use in digital forensics work, the first problem seems to be transferring them from their device to my computer. It is a familiar task, but in this context receiving the files is only part of the work: I also need to document what I received and give whoever examines them after me a way to verify their integrity.
Months later, another consultant might need to check those files without having witnessed the handover and without being able to ask the client to repeat it. The way I acquire them therefore needs to leave enough of a record to reconstruct that step, rather than relying on my account of it.
That need led me to develop Quickshare, an open source tool for receiving files from a device without installing anything on it. The computer creates a dedicated Wi-Fi network; the client connects, opens a page using a QR code, and selects the files to send through the browser. The interaction is simple, but one question guided the design choices behind it: what records and checks will someone need to verify this acquisition?
Identifying what we received
A file’s name is not enough to identify it. Two different photographs can both be called IMG_0001.jpg, and a document can keep the same name after it has been modified. To verify that the file we are examining is identical to the one received, we need a reference tied to its contents.
A hash function such as SHA-256 lets us calculate a file’s fingerprint: a sequence of fixed length that we can recalculate at any time. If the contents change, the fingerprint changes; obtaining the same fingerprint from two different sets of contents is considered impractical. Keeping this value lets us compare a file with the reference recorded at the time of acquisition.
This check has a specific meaning: it tells us whether the contents match those for which we calculated the fingerprint. It does not establish who created the file, whether a photograph really depicts what is claimed, or whether a document had already been modified before the handover. Those are different questions that require other evidence and analysis.
Where verification begins
Calculating the fingerprint only after receiving a file leaves a question unanswered: does what arrived on the computer match what was selected on the source device? We would have a reference for later checks, but no comparison between the contents before and after transfer.
That is why Quickshare calculates the fingerprint on both devices. The browser calculates it from the file selected on the source device, and the computer calculates it during reception. Comparing the two lets us verify that the transfer preserved the contents. The client also sees the result on their own screen.
The starting reference for verification is therefore the file selected for upload. From that point, we can check its transfer from the source device to the computer and keep a fingerprint against which to compare subsequent copies.
Documenting the handover
The fingerprint identifies the contents, but on its own tells us little about the acquisition. To reconstruct the handover, we also need information about when it happened, which device the file came from, and the name under which it was stored.
Quickshare accompanies each batch of received files with a log recording these details alongside file sizes and fingerprints. The log connects what we find in the folder with the operations performed during the handover and provides a record of that stage in the chain of custody.
The log itself is also a file, though, and could be modified. When it is closed, a fingerprint of its contents is calculated, to be recorded in a separate document as well, such as the acquisition report. This step matters: if the reference existed only in the log, someone modifying it could also recalculate the fingerprint. Keeping it elsewhere allows the log to be checked against an external reference.
Verification also extends to the tool used. The log therefore records Quickshare’s version, the commit identified at build time, and the executable’s fingerprint. These references help identify the program used and the version of the code to examine. Publishing the source also allows another consultant to study how it works, as well as check the files it produced.
When something goes wrong
Documenting an acquisition also means accounting for things that can go wrong during the handover. Two files might have the same name, a connection might drop, or the client might close the page before an upload finishes.
If one file overwrote another, we would lose some of the material received. If an interrupted transfer left an incomplete file in the folder, we might mistake it for a successful acquisition. Quickshare keeps files with identical names separate, removes incomplete files, and logs interrupted uploads, allowing us to distinguish what arrived from what was only attempted.
Restricting access to the session serves the same purpose of defining the boundaries of the handover we are documenting. The dedicated network and session-specific link help control who can submit material; access to the service is restricted to the network used for the handover and to the duration of the operation.
For whoever comes next
Once the acquisition is complete, whoever receives the material needs to be able to connect the files to the log, compare their fingerprints, and identify the tool used. Each record or check answers a concrete question and makes part of the work verifiable by someone who was not there.
This is the principle I try to apply in digital forensics: documenting the steps so another professional can examine and assess them. A transfer takes a few minutes, but the records it leaves need to remain useful even after the details of that handover have been forgotten.